torch.Tensor和numpy.ndarray

shaoheshaohe

1. torch.Tensor和numpy.ndarray相互转换

[url=][/url]
import torchimport numpy as np# <class 'numpy.ndarray'>np_data = np.arange(6).reshape((2,3))# <class 'torch.Tensor'>torch_data = torch.from_numpy(np_data)# <class 'numpy.ndarray'>tensor2array = torch_data.numpy()print('numpy array:\n',np_data,type(np_data),      '\ntorch tensor:\n',torch_data,type(torch_data),      '\ntensor to array:\n',tensor2array,type(tensor2array))

shaoheshaohe

3. 矩阵乘法(正确的做法)
data = [[1,2], [3,4]]
tensor = torch.FloatTensor(data)
print(
    '\nmatrix multiplication (matmul):',
    '\nnumpy:\n', np.matmul(data, data),     # [[7, 10], [15, 22]]
    '\ntorch:\n', torch.mm(tensor, tensor))  # [[7, 10], [15, 22]]

shaoheshaohe

torch.Tensor:是一个包含了一种数据类型元素的多维矩阵，缺省为torch.FloatTensor
2. torch.Tensor和numpy.ndarray一些简单操作，如均值，绝对值，sin，log等
复制代码
data = [-1,-2,1,2]
tensor_default = torch.Tensor(data)
tensor = torch.FloatTensor(data)
print('tensor default type:\n',tensor_default,
      '\ntensor FloatTensor type:\n',tensor,
      '\nabs:',
      '\nnumpy:',np.abs(data),
      '\ntorch:',torch.abs(tensor),
      '\nsin:',
      '\nnumpy:',np.sin(data),
      '\ntorch:',torch.sin(tensor),
      '\nmean:',
      '\nnumpy:',np.mean(data),
      '\ntorch:',torch.mean(tensor),)